Are virtual physiology laboratories effective for student learning? A systematic review

Abstract

It is unclear if the transition from traditional, in-person physiology laboratories to virtual alternatives has educational impacts on students. This study used a systematic review to critically evaluate research papers that investigated the effectiveness of virtual physiology laboratories for student learning. Eleven studies, retrieved from the Education Resources Information Center (ERIC) and Ovid MEDLINE databases, were selected for inclusion in this review, based on predetermined eligibility criteria. Subsequently, the studies went through a power analysis for potential biases before their results were synthesized and analyzed. This systematic review found that virtual physiology laboratories are effective for students’ learning of concepts. However, it was inconclusive as to whether virtual physiology laboratories are effective for students’ motivation for learning and learning of technical skills. It was found that blended models of virtual laboratories are at least as effective as in-person laboratories for conceptual learning. Overall, this systematic review provides useful insights for educators regarding the educational impacts of implementing virtual laboratories into the physiology curriculum and suggests research models for future evaluation of virtual laboratories.

INTRODUCTION

The Problem: A Forced Shift to Virtual Laboratories

In 2020, the coronavirus disease 2019 (COVID-19), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), led to a pandemic. In response to this pandemic, the Australian Government declared a human biosecurity emergency and imposed lockdown measures to contain the transmission of the virus (1). This included social distancing, one of the most effective strategies to limit the spread of the virus through human-to-human transmission (2). To comply with these social distancing measures, the delivery of in-person teaching models at universities, including laboratories, was forced to shift to remote, or virtual, alternatives. Laboratories, a critical component of undergraduate science education (3, 4), have thus been severely impacted by the COVID-induced restrictions (5, 6). For the physiology discipline, traditional, in-person laboratories play a central role to support students’ understanding of physiology concepts and the development of research and technical skills (7–9). In particular, the social interactions and the teamwork during in-person laboratories support collaborative learning (3). As a result of university lockdowns, the remote delivery of virtual physiology laboratories is thus likely to have educational impacts on student learning.

Historical Development of Virtual Laboratories

A virtual laboratory is any online environment that is based on interactive learning either individually or in groups, allowing students to explore topics in an asynchronous manner that has no immediate physical reality (10). Virtual laboratories have been developed since the 1990s, where educators used readily accessible technologies such as the compact disk read-only memory (CD-ROM) to design virtual laboratories that could be accessed by students (11). Contemporary designs of virtual laboratories utilize network-based configurations to create virtual simulations for manipulation. In these virtual laboratories, participants manipulate virtual apparatus to perform experiments in computer-generated simulations. For example, virtual reality computer programs have been used to teach robotics to engineering students (12). Additionally, web packages, which are website-based platforms with interactive components, have been used to teach undergraduate students about the cardiovascular system (13). Furthermore, virtual simulations of animal experiments have been used for teaching physiology (14–16). In the last decade, there has been a gradual shift of conventional physical laboratories toward virtual alternatives, to either fully replace or support the physical laboratories (17). There is also an increasing number of university disciplines teaching via web-based instruments and more university courses that are being delivered entirely online (18).

Pre-COVID Reasons for the Use of Virtual Laboratories

Before the COVID-19 pandemic, there were four main reasons that promoted the transition of physical, in-person, laboratories to virtual alternatives. First, conducting physical laboratories is financially demanding. This can be attributed to the provision of advanced instruments, space, the recruitment of professional personnel, the maintenance of equipment, and the rapidly increasing student population (2, 19). Second, animal-based laboratories can be associated with ethical concerns, therefore, limiting the scope of experiments that educators can employ for students (20). In comparison, using virtual animal model simulations reduces the ethical dilemmas and broadens the types of experiments that can be conducted. Moreover, as compared with in-person laboratories, virtual laboratories exhibit higher levels of efficiency (21) and safety (22, 23). Lastly, prior studies have reported that virtual learning strategies such as simulations and instructional videos can be used to support the physical laboratories in enhancing students’ learning (7, 24–26). For these reasons, the transition from traditional in-person laboratories to virtual alternatives is likely to persist in the higher education sector. However, in-person laboratories, which provide tactile hands-on experiences, are undeniably different in nature to the virtual laboratories, which have no physical reality. Therefore, this transition is likely to impact student learning. While the in-person laboratories are unquestionably effective for student learning (27), it is unclear as to whether virtual laboratories support student learning.

Student Learning for Physiology Laboratories

Due to the increasing use of virtual laboratories, particularly in response to the COVID-19 pandemic, it is critical to determine if they are effective for student learning. For this study, effective student learning was defined as student achievement of the (laboratory) learning outcomes (28). Laboratory learning outcomes cover both physiology content/concepts and skills or competencies. The type of student learning from laboratories was divided into three categories (29):

1. Conceptual learning: the laboratory reinforces or enhances the understanding of physiology concepts taught in other teaching modes, such as lectures and workshops;

2. Motivational learning: the laboratory motivates student engagement with scientific processes, scientific exploration and active learning from failures (in experimental design, for example); and

3. Technical learning or research skills development: the laboratory allows the practice and learning of technical skills such as using scientific equipment and other research skills such as conducting data analysis, interpretation, and presentation.

Systematic Review for Evaluating the Research Literature

In the physiology discipline, despite a number of primary research articles about virtual laboratories (for example, see Refs. 14, 24, 30), there has been no critical evaluation of these papers. This study used a systematic review to critically appraise the existing literature, to determine if virtual physiology laboratories are effective for student learning. A systematic review is a robust process that critically analyses previous literature through explicit methodology, minimizing biases and enabling the synthesis and generation of comprehensive conclusions (31–33). The systematic review was expected to identify articles that highlight the educational impacts of virtual physiology laboratories on student learning and to reveal gaps in the research literature, guiding the direction of future research.

METHODS

Study Design

This systematic review was conducted following the Preferred Reporting Items for Systematic Reviews & Meta-Analyses (PRISMA) guidelines (34), a framework that allows the design of well-defined methodology to evaluate the existing literature from an unbiased and objective view (31).

Search Strategy

The electronic database search included the Education Resources Information Center (ERIC) and Ovid MEDLINE. ERIC was selected because of its extensive collection of educational research that is expected to include a large number of research publications for this systematic review. Ovid MEDLINE was selected because it is an authoritative database that includes a comprehensive collection of medical-related research that is relevant to physiology. The scope of the literature search was limited by five categories, including 1) virtual; 2) laboratory; 3) undergraduate students; 4) physiology discipline(s); and 5) educational benefits. Keywords and subject headings from each category were used to conduct the literature search (see Table 1).

Table 1.

Keywords and subject headings for the literature search

Categories	Keywords†	Subject Headings (Ovid MEDLINE)	Subject Headings (ERIC)
Virtual	Virtual Simulat* Blended learning Blended teaching Computer* Online	Computer simulation (methods, standards) Virtual reality Simulation training (methods, standards) Online systems (education, instrumentation methods, standards)	Virtual classrooms Virtual universities Computer uses in education Computer simulation Online systems Simulation
Laboratory	Lab* based Lab* class* Lab* setting* Practical class* Practical lab* Experiment*	Laboratories (instrumentation, methods, standards)	Laboratory techniques Laboratory training Laboratory equipment
Undergraduate students	Undergraduate*	Universities (education, standards) Education, medical, undergraduate (methods, standards) Teaching (education, organization and administration, standards, statistics and numerical data)	College students Undergraduate students
Physiology discipline‡	Physiology Animal* Neurophysiology Endocrinology Immunology Pathophysiology Psychophysiology	Physiology (analysis, classification, drug effects, education, ethics, instrumentation, methods, standards, organization and administration, statistics and numerical data, supply and distribution, surgery, trends) Animal testing alternatives (education, ethics, instrumentation, methods, standards)	Exercise physiology Exercise (physiology) Hearing (physiology) Psychophysiology Animals Physiology
Educational benefits	Benefit* Advantage* Effective* Education* Pitfall* Limitation* Disadvantage* Evaluat*	Evaluation study Program evaluation (methods, standards, statistics and numerical data, trends)	Educational opportunities Educational assessment Educational benefits Program evaluation

^†Same keywords were used in both Ovid MEDLINE and Education Resources Information Center (ERIC) during the literature search, whereas the subject headings were restricted to the unique thesaurus of each database, therefore, suitable subject headings were selected from each database. The asterisks (*) allow the retrieval of all words that have the same starting letters and variable ending letters. ‡Including the keyword “physiology” will retrieve all terms including these letters; therefore, subdisciplines such as “respiratory physiology” were not included as keywords.

The search was conducted on the 21st of September 2020, using the following steps: 1) for each category, the keywords were searched individually; 2) for each category, the subject headings were searched individually; 3) for each category, the search results of all keywords and subject headings were combined with the operator “OR”; 4) the search results from all categories were combined with the operator “AND”; and 5) the search was limited to papers from 2005 as this aligned with the period for significant development of virtual laboratory technologies.

Eligibility Criteria

Database search results were imported into an EndNote library, including the portable document format (pdf) of the full-text articles. Duplicates were then removed, and initial screening was completed by title and abstract. This was followed by screening of the full-text articles. Exclusion of papers was completed by the three researchers independently, followed by consensus for exclusion by the researchers. The relevant inclusion criteria are identified below, and reasons for exclusions are noted in the PRISMA flow chart (Fig. 1):

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart of study selection process. The process followed the sequence of removing duplicates, screening the titles and abstracts, screening the full-texts, and including relevant studies. The reasons for exclusion of the full texts are included in this schematic.

1. Language: the literature was presented using the English language;

2. Full text;

3. Peer-reviewed articles;

4. Physiology discipline: the function of organisms, including subdisciplines such as cell physiology, neurophysiology, endocrinology, immunology, exercise physiology, pathophysiology, comparative physiology, cardiovascular physiology, respiratory physiology, psychophysiology, renal physiology, and systems physiology;

5. Undergraduate: tertiary/undergraduate education, after secondary education and before postgraduate education; and

6. Critical evaluation of virtual physiology laboratories: the literature conducted formal evaluation(s) of the virtual instrument used in the study from any of the following aspects: conceptual learning, motivational learning, and technical learning.

A Google Excel spreadsheet was created to systematically screen the literature. The spreadsheet included the basic information (title, citation), the key findings, and the eligibility criteria to select the relevant papers. The screening process began by checking the titles and the abstracts of the articles against the eligibility criteria. Following the elimination of irrelevant articles by titles and abstracts, the full texts of the remaining papers were assessed against the eligibility criteria. After the screening process, a team consensus was reached for the eligibility criteria, before finalizing the relevant papers for analysis.

Assessment of the Risk of Bias

As the included studies comprised a mixture of both quantitative and qualitative studies, the conventional data analysis tools were not applicable to this review. Therefore, the risk of bias in individual studies was evaluated by using a novel additive scoring system. Firstly, nine bias/error-assessing criteria were nominated to evaluate the risk of bias/error of the studies. This included 1) randomization; 2) sample size; 3) voluntary participation; 4) presence of control group(s); 5) group sizes; 6) consistency across groups; 7) response rate; 8) type of evaluation; and 9) data significance (see Table 2). A score was allocated to each criterion, and then the scores were added for all of the criteria to produce a final score (out of 5) that indicates the power of the study. Criteria 1, 4, 5, and 6 were considered together as a lack of control group(s) simultaneously invalidates the other three criteria. Criteria 3 and 7 were considered together as although volunteerism can increase the risk of volunteer bias, a high response rate suggests a high level of inclusion of the sample with a lower risk of sampling bias; therefore, each criterion was allocated 0.5 points. Evaluation of the literature involved examining the fulfilment of the criteria, and the respective points were scored. The threshold of valid sample sizes used in this study was 50 (35), and the threshold of valid response rate used in this study was 60% (36). Table 2 summarizes the criteria, their definitions, and their score allocations.

Table 2.

Method for the assessment of potential bias/error within studies

No	Criteria	Definition	Evaluation (Score Allocation)
1	Randomization	The random allocation of participants to the experimental/control groups.	Yes (0.25): The presence of randomization indicates a lower risk of selection bias. No (0): The absence of randomization indicates a higher risk of selection bias. N/A (0): Not applicable for studies with only 1 group.
2	Sample size	A minimum of 50 participants are required to establish a valid evaluation.	Large (1): Large sample size indicates a lower risk of sampling bias. Small (0): Small sample size indicates a higher risk of sampling bias.
3	Voluntary participation	The presence of volunteerism in the study, which includes volunteering to participate in the study or provide data for the study.	Yes (0): Using volunteers for a study indicates a higher risk of volunteer bias. No (0.5): Using nonvolunteers for a study indicates a lower risk of volunteer bias.
4	Presence of control group(s)	A group without the implementation of the target intervention. It is compared with the experimental group, where the target intervention is implemented.	Yes (0.25): The presence of control group(s) indicates a lower risk of placebo effect. No (0): The absence of control group(s) indicates a higher risk of placebo effect.
5	Group sizes	The number of participants per group is balanced.	Yes (0.25): Balanced group sizes indicate a lower risk of selection bias, sampling bias, and type I errors. No (0): Unbalanced group sizes indicate a higher risk of selection bias, sampling bias, and type I errors. N/A (0): Not applicable for studies with only 1 group.
6	Consistency across groups	Demographics of participants across the groups are consistent without significant variation between them.	Yes (0.25): Consistent demographics across groups indicates a lower risk of selection bias and sampling bias. No (0): Inconsistent demographics across groups indicates a higher risk of selection bias and sampling bias. Not evident (0): The risks of biases cannot be determined. N/A (0): Not applicable for studies with only 1 group.
7	Response rate	The proportion of participant response compared with the total number of participants in the cohort/subject.	High (0.5): High response rate indicates a lower risk of nonresponse bias. Low (0): Low response rate indicates a higher risk of nonresponse bias.
8	Type of evaluation	Survey/Questionnaire vs. Test score.	Survey/Questionnaire (0.5): Evaluation using surveys/questionnaires can be subjected to response biases. Test score (1): Evaluation using test score eliminates response biases.
9	Data significance	The significance of data collected is investigated.	Yes (1): Investigating the data significance indicates a lower risk of interpretive bias. No (0): Not investigating the data significance indicates a higher risk of interpretive bias.

Data Items

The mode of delivery is defined as the educational configuration of the laboratory activity. For example, a blended mode of delivery involves both the components of the physical laboratory as well as the components of the virtual laboratory. In contrast, a standalone virtual mode of delivery includes no physical laboratory. A critical evaluation of virtual tools refers to the formal assessment of the virtual instrument used in the study for student learning from any of the following aspects: conceptual learning, motivational learning, and technical learning.

Ethics Statement

Institutional Review Board approval was not necessary for this protocol as it was a systematic review, using data form publicly accessible data.

RESULTS

Literature Search

The literature search retrieved 57 papers from ERIC and 58 papers from Ovid MEDLINE, summing up to 115 papers in total (Fig. 1). After removal of 11 duplicates, the abstracts for the remaining 104 papers were screened against the eligibility criteria, following the order of English, full-text, primary journals, physiology discipline(s), undergraduate level, virtual laboratory, critical evaluation. All papers were in English and were primary studies. This resulted in the exclusion of 75 papers. When the full text for the remaining 29 papers was assessed, 16 papers were excluded (see Fig. 1 for the breakdown of the exclusion criteria). A total of 13 papers were included for subsequent evaluation. Detailed information about the number of papers excluded and the corresponding reasons for exclusion at each stage of screening are shown in Fig. 1.

Risk of bias assessment.

Based on the novel additive scoring system used, 2 out of the 13 included studies (papers) had low power and high risk of bias/error. A revisit of these studies revealed deficiencies in their study designs; therefore, both studies were removed from this review.

Data Extraction from the Included Studies

The participants, study design, research aims, laboratory, and key findings for the final 11 included studies in this systematic review are presented in Table 3.

Table 3.

Study characteristics

Reference	Participants	Study Design	Research Aims	Laboratory	Key Findings
Moreno-Ger et al. (37)	143 medical undergraduates	Randomized control trial design Experimental group (EG), n = 66, used the virtual laboratory simulation prior to the physical laboratory Control group (CG), n = 77, attended the physical laboratory only Student perception: 5-point Likert survey	To investigate: If having prelab virtual simulation decreases the perceived difficulty for the physical laboratory. If having prelab virtual simulation increases the experimental precision during physical laboratory. The students’ attitudes toward a prelab virtual simulation.	Aim: To develop students' skills to quantify different parameters from a blood sample taken from an experimental animal. Virtual tool: Virtual laboratory simulation Mode of delivery:*Blended Asynchronous	The EG perceived that the procedures were easier to understand and displayed a higher level of precision in determining the hematocrit. The EG reported a more positive experience.
Motz et al. (38)	173 undergraduates from the Department of Psychological and Brain Sciences	Quasiexperimental study design EG, n = 104, CG, n = 69, Pretest and posttest results were compared.	To investigate if the computer program improves students’ conceptual understanding of neurophysiology.	Aim: To discover the lateralization of cognitive function, the anatomy of the brain, and the spatial frequency hypothesis. Virtual tool: Computer program, Lateralizer. Mode of delivery: Virtual Asynchronous	EG made stronger improvements between the pretest and posttest than CG. However, due to significant inconsistency between CG and EG, the comparison between CG and EG was considered invalid. Nonetheless, the improvement within the EG was significant.
Mutlu et al. (39)	71 undergraduates from a nursing faculty	Randomized control trial design High-fidelity group (HFS), n = 36, used Nasco Smartscope Simulator Low-fidelity group (LFS), n = 35, used computer-based simulator Pretest and posttest results were compared.	To determine the effects of high- and low-fidelity simulators on student nurses’ learning of heart and lung sounds.	Aim: To develop students’ skills to accurately identify heart and lung sounds. Virtual tool: Low-fidelity simulator (computer-based simulator)† Mode of delivery:Virtual Synchronous	The high-fidelity simulator was more effective compared with the low-fidelity simulator in supporting students’ conceptual understanding in heart and lung sounds.
Wang et al. (16)	63 undergraduates (Bachelor of Medicine and Bachelor of Surgery)	Randomized control trial design Living tissue laboratory group, n = 23 Virtual laboratory group, n = 18 Blended group, n = 22 Median postlaboratory quiz score (points) compared between groups Student perception survey (4-point Likert-scale format).	To investigate the impact of the virtual laboratory simulation in supporting students’ learning and perceptions of neurophysiology. To assess the validity of the virtual laboratory simulation as a useful pedagogical tool for neurophysiology.	Aim: To enhance students’ understanding on the generation and conduction of neural action potentials. To provide basic training of measuring neuronal excitability and signal conduction.Virtual tool: Virtual laboratory simulation Mode of delivery: Virtual and blended Synchronous	Both the live tissue laboratory and the blended laboratory were more effective in supporting conceptual learning than the virtual laboratory. The live tissue, blended, and virtual laboratories were equally effective in supporting students’ technical learning. Delivering animal experiments via virtual means was deemed less attractive compared with the physical and the blended delivery modes.
Chen et al. (24)	173 medical undergraduates 34 teachers who taught the blended laboratory course	Within-subject design Students undertook the traditional laboratory course in 2015 and a blended laboratory course in 2016. Average scores for laboratory quizzes were compared. Student and teacher perception surveys were conducted.	To investigate if the blended laboratory course supports students’ understanding in threshold concepts and improves their abilities of self-learning, understanding, and problem solving.	Aim: Not applicable as the study evaluated multiple laboratories. There is no explicit description of the laboratories. Virtual tool: Blended laboratory course, including micro videos Mode of delivery:Blended Asynchronous	The blended laboratory course was effective in enhancing test scores. Students preferred the blended laboratory course compared with the traditional methods. Teachers had ascertained the effectiveness of the blended model of virtual physiology laboratories in supporting students’ conceptual and motivational learning.
Dantas and Kemm (7)	31 undergraduates studying a physiology course in a Science degree	Within-subject design Students completed e-Learning activities as prelab exercises. Student perception survey (5-point Likert-scale). Correlation between e-learning marks and final examination marks was analyzed.	To investigate student use of e-learning in the laboratory-based course.To evaluate the learning outcomes that are enhanced by active learning facilitated by the e-learning component of the course.	Aim: Not applicable as the study evaluated multiple laboratories. There is no explicit description of the laboratories. Virtual tool: e-Learning platform, with active learning promoted through hypothesis testing and predictions for physical laboratories. Mode of delivery: Blended Asynchronous	The majority of the students did not agree that the virtual physiology laboratory motivated them to learn. The e-learning marks correlated with final examination masks.
Dobson (40)	25 undergraduates from the Department of Applied Physiology and Kinesiology	Crossover study design. EG 1, n = 12, completed the traditional exercise physiology laboratory before undertaking a knowledge assessment. Next, the group completed the Virtual Physiology of Exercise Laboratory (VPEL) program before undertaking a knowledge assessment. EG 2, n = 13, completed the VPEL program before undertaking a knowledge assessment. Next, the group completed the traditional exercise physiology laboratory before undertaking a knowledge assessment.	To compare student learning (experimental setup, conduct, and data analysis) from the VPEL program with that from traditional exercise physiology laboratory activities. To determine if the order in which the different types of laboratories were delivered affects students’ conceptual understanding in exercise physiology.	Aim: To enhance students’ conceptual understanding in lactate and ventilatory threshold and maximal oxygen consumption during exercise. To allow students to practice the processes of setting-up and conducting experiments, and analyzing the data collected. Virtual tool: VPEL simulation Mode of delivery: Virtual Synchronous	No significant differences between the 2 groups for experimental assessment.
Durand et al. (41)	350 medical undergraduates	Cross-sectional study design Group A, n = 120, conducted animal laboratory classes. Group B, n = 122, conducted both animal and virtual laboratory classes. Group V, n = 108 conducted virtual laboratory classes. Student perception survey was carried out.	To compare the students’ perceptions of animal laboratory classes versus video/computer-based learning in physiological sciences associated with the effectiveness of the PBL method.	Aim: Not applicable as the study evaluated multiple laboratories. Virtual tool: Online videos with interactive activities Mode of delivery: Virtual‡ Synchronous	Both the animal-using laboratories and the virtual physiology laboratories were found to support students’ motivational and conceptual learning. Students viewed the use of animals as more effective to stimulate their motivational learning. Students agreed that both the animal-using laboratories and the virtual physiology laboratories were effective to support technical learning, however, they differed in terms of the skill sets that they developed.
Elmer et al. (25)	33 undergraduates from the Department of Kinesiology and Integrative Physiology	Crossover study design Section 1, n = 16, undertook the laboratories in the order of: blended neuromuscular power laboratory, traditional maximal oxygen consumption laboratory, blended blood lactate laboratory, traditional muscle electromyography (EMG) laboratory. Section 2, n = 17, undertook the laboratories in the order of: traditional neuromuscular power laboratory, blended maximal oxygen consumption laboratory, traditional blood lactate laboratory, blended muscle EMG laboratory. Student perception survey was carried out.	To compare student performance and perceptions between blended and traditional laboratories in an undergraduate exercise physiology.	Aim: Not applicable as the study evaluated multiple laboratories. There is no explicit description of the laboratories. Virtual tool: Video demonstrations and knowledge checks Mode of delivery:Blended Synchronous	The blended model of virtual physiology laboratory was equally effective as the traditional living tissue experiments in improving students’ performances. The majority of the students agreed that the virtual component of the blended model greatly enhanced their conceptual learning and engaged them to perform physiology experiments. The majority of the students preferred the blended model to the traditional model.
Gopal et al. (13)	165 undergraduates undertaking the Anatomy and Physiology course	Quasiexperimental study design EG, n = 80, had access to the virtual learning platform CG, n = 85, did not have access to the virtual learning platform Students’ lab tests scores and their usage of the virtual platform were analyzed.	To compare student performance among undergraduate students with access to a virtual learning platform to students without access to the platform.	Aim: Not applicable as the study evaluated multiple laboratories. There is no explicit description of the laboratories. Virtual tool: Web-based cardiovascular system platform Mode of delivery: Blended Synchronous? (this is unclear)	Students with access to the virtual platform took advantage of the virtual platform, with EG showing higher performance than CG.
Modica et al. (42)	227 medical undergraduates	Quasiexperimental study design EG, n = 120, completed the experimental curriculum, including web-based musculoskeletal tutorial, pathophysiology-focused cases, and facilitator preparation. CG, n = 107, completed a traditional musculoskeletal curriculum. Multiple-choice examination and student perception survey were carried out.	To assess the effectiveness of the experimental curriculum on teaching students the musculoskeletal exam as compared with a traditional curriculum. To assess students’ and preceptors’ satisfaction when exposed to our experimental teaching method.	Aim: To enhance students’ conceptual understanding and technical skills in musculoskeletal examination. Virtual tool: A web-based curriculum, including musculoskeletal tutorial, pathophysiology-focused cases, and facilitator preparation. Mode of delivery: Blended Asynchronous (web-based tool) and synchronous (practice session)	No significant differences between the two groups for the examination. Students viewed the PFCs (virtual laboratory component) as the most useful in supporting their conceptual learning. The additional resources offered through the blended model of virtual physiology laboratory made the blended model the preferred mode of delivery to traditional laboratories for students.

EG, experimental group; CG, control group. *Mode of delivery refers to the form of learning undertaken by the studies in offering the virtual instruments. †Only low-fidelity simulator is applicable, as the high-fidelity simulator (nasco smartscope simulator) is a mannikin simulator that can only be used in physical laboratory settings. ‡Blended is not included as group B experienced a transition from having only animal laboratories to having only virtual laboratories.

Demographics.

All of the final included studies for this systematic review were conducted at different universities. A large proportion of the studies were from the United States (45.5%), followed by China (18.2%), Australia (9.1%), Brazil (9.1%), Spain (9.1%), and Turkey (9.1%). The participants of the studies were largely undergraduate students including medical (45.5%), exercise physiology (18.2%), and nursing students (9.1%). The remaining participants were from science degrees (27.3%), including biomedical and biological science.

Study designs.

Three of the eleven studies conducted randomized controlled trials, three adopted a quasiexperimental study design, two adopted a within-subject design, two adopted a crossover study design, and one adopted a cross-sectional study design (see Table 3).

Virtual tools and modes of delivery.

The virtual tools and modes of delivery for the 11 papers are summarized in Table 3 For the virtual laboratories used, 36.4% were virtual simulations (usually an imitation of the physical laboratory), 27.3% were web-based laboratory courses (website with interactive activities), 27.3% were video demonstrations (video displaying experimental procedures and outcomes), and 9.1% were computer software (inquiry-based system that allows students to perform experiments). Of the 11 studies, 5 were delivered in an asynchronous manner (7, 24, 37, 38, 42), 5 were delivered in a synchronous manner (13, 16, 39–41), and 1 was delivered via both modes (25). To test these virtual laboratories, 63.6% of the studies conducted the physiology laboratories using blended models and 45.5% of the studies conducted the physiology laboratories using standalone virtual models. The study from Wang et al. (16) adopted both the blended and the standalone virtual models of physiology laboratories.

Synthesis of Results

The effectiveness of the virtual physiology laboratories for students’ conceptual, motivational, and technical learning was defined as an achievement of the laboratory aims and learning outcomes, as shown in Table 3. Seven studies explored the potential of virtual laboratories for supplementing the in-person physiology laboratories, with a blended model (see Table 3). Six of these seven studies showed that the blended model of delivery was more beneficial compared with the in-person delivery, by providing additional benefits such as educational resources.

Are virtual laboratories effective for conceptual learning?

A breakdown of the effectiveness of virtual laboratories for conceptual learning is summarized in Fig. 2. Ninety percent of the studies that investigated conceptual learning found that virtual laboratories supported this, with most of these studies using a blended model for the virtual laboratories.

Figure 2.

Efficacy of virtual physiology laboratories for student’s conceptual learning. Ten out of the eleven studies included in the systematic review of virtual physiology laboratories investigated conceptual learning [Motz et al. (38); Durand et al. (41); Dobson et al. (40); Chen et al. (24); Dantas and Kemm (7); Elmer et al. (25); Gopal et al. (13); Modica et al. (42); Wang et al. (16); Mutlu et al. (39)]. Nine of these ten studies found that virtual physiology laboratories were effective for conceptual learning, with most of these studies using a blended model for the delivery of the virtual laboratories. One study investigated standard virtual elements in laboratory simulations compared with advanced virtual elements (i.e., virtual reality) for conceptual learning.

Are virtual laboratories effective for motivational learning?

Six out of the eleven studies investigated whether the virtual physiology laboratories were effective for motivational learning. Five out of the six studies (83.3%) supported this, with three of these studies comparing blended models with in-person models (24, 25, 37), one study comparing virtual with in-person laboratories (41), and one study comparing both virtual and blended models with an in-person model (16). One study compared virtual models with in-person models and showed that the virtual models were not effective for students’ motivational learning (7).

Are virtual laboratories effective for technical learning?

Three out of the eleven studies investigated whether the virtual physiology laboratories were effective for technical learning. All three studies (100.0%) supported this, with one study comparing a blended model with an in-person model (37), one study comparing virtual models with in-person models (41), and one study comparing both virtual and blended models with an in-person model (16).

Educators’ personal experiences of virtual laboratories.

Researchers from three studies provided their personal opinions regarding the replacement of in-person physiology laboratories with virtual alternatives (13, 16, 40). Although not experimentally tested, the researchers emphasized that the experience provided by the in-person laboratories cannot be replaced by virtual means. However, with further development to create more holistic learning experiences, the virtual laboratories exhibit the potential to replace the conventional laboratories in the future.

DISCUSSION

This study used a systematic review to determine if virtual physiology laboratories are effective for student learning. Analysis of the 11 included studies found that virtual laboratories supported conceptual, motivational or technical learning. However, there were limitations with the mode of delivery (63.6% of the studies used blended models) and the number of studies, which will be covered in this discussion.

Conceptual Learning

The majority of the studies that investigated conceptual learning showed that the virtual physiology laboratories were effective for students’ conceptual learning. This was tested using quantitative analysis of test scores, with the “pretest-posttest” design (13, 24, 38, 39) or the “posttest-only” design (7, 16, 25, 40–42). The pretest-posttest design allows the measurement of baseline values to be used as references to show the significance of changes before and after the intervention (43), whereas the posttest-only design measures only the performance of the participants after the intervention without baseline values, which can induce biases and errors (44). Therefore, the studies that utilized the pretest-posttest design showed a higher level of validity and reliability as compared with the studies that utilized the posttest-only design.

It is likely that the mode of delivery of the virtual laboratory will also have impacts on conceptual learning. Three studies compared physiology laboratories using standalone virtual models and in-person models (see Fig. 2). Specifically, the study by Dobson (40) compared an in-person exercise physiology laboratory with a virtual simulation through a Virtual Physiology of Exercise Laboratory (VPEL) program. The study by Durand et al. (41) compared students’ perceptions and conceptual understanding of physiology laboratories that used animals and simulated physiology laboratories. In conclusion, both studies reported that the standalone virtual physiology laboratories were as effective (for conceptual learning) as in-person physiology laboratories. On the other hand, the study by Motz et al. (38), which compared a computer-based neurophysiology laboratory with an in-person neurophysiology laboratory, had inconsistent experimental and control groups, which invalidated the comparison between them. Nonetheless, their analysis of the pre- and posttest scores of only the experimental (standalone virtual laboratory) group revealed that the standalone virtual laboratory was effective for students’ conceptual learning.

The blended model of virtual physiology laboratories was found to exhibit an equal, if not a higher, level of effectiveness for students’ conceptual learning than in-person laboratories. Among the five studies that compared physiology laboratories in the blended and in-person contexts, the studies by Chen et al. (24) Dantas and Kemm (7), Elmer et al. (25), and Gopal et al. (13) compared multiple physiology laboratories in their virtual and blended models, whereas the study by Modica et al. (42) compared only the muscle physiology laboratories with virtual and blended models. The studies by Chen et al. (24) and Gopal et al. (13) reported that their blended models of physiology laboratories were more effective than the in-person laboratories for conceptual learning, while the studies by Dantas and Kemm (7), Elmer et al. (25), and Modica et al. (42) showed that their blended models of physiology laboratories were as effective as the in-person physiology laboratories.

Interestingly, the study by Wang et al. (16), which compared both a standalone virtual model and a blended model of neurophysiology laboratories (virtual and in-person) against an in-person neurophysiology laboratory involving live tissue dissections, showed that while the blended model was as effective as the traditional laboratory, the standalone virtual laboratory was less effective than the in-person laboratory. Hence, it showed that a blended laboratory curriculum with both virtual and physical components exhibits greater educational benefits than a standalone virtual model.

The study by Mutlu et al. (39) did not support the efficacy of virtual physiology laboratories for students’ conceptual learning. They compared an advanced technology (mannikin simulator), which was embedded into an in-person laboratory, with a less technically advanced virtual simulation (computer simulation) of the physiology laboratory. Therefore, although the virtual simulation was deemed less effective than the mannikin simulator, this supported the fact that students’ conceptual learning is effectively supported by implementing advanced technology into in-person physiology laboratories.

Therefore, for this systematic review, virtual physiology laboratories were found to be effective for students’ conceptual learning, with evidence that the blended model is an equal, if not a more beneficial, mode of physiology laboratory, to virtual laboratories.

Motivational Learning

Most of the studies that investigated motivational learning reported that virtual physiology laboratories supported this, with the majority of the students expressing that the virtual forms of physiology laboratories motivated them to actively explore physiology concepts. They evaluated motivational learning using qualitative analysis (subjective feedback) based on questionnaires and surveys for students (16, 24, 25, 37, 41) and educators (24).

The mode of delivery of the virtual laboratory could impact motivational learning. The study by Durand et al. reported that the standalone virtual physiology laboratories were effective for motivational learning by stimulating students’ interest, however, this was to a lesser extent than the in-person laboratories. The study by Moreno-Ger et al. (37), which investigated the educational impacts of a prelaboratory virtual simulation for a cardiovascular physiology laboratory, reported that the blended model of physiology laboratory was effective in stimulating students’ motivational learning by presenting the concepts in forms that are easier to understand. Similarly, Chen et al. (24) and Elmer et al. (25) reported that the blended models of physiology laboratories were more effective for students’ motivational learning than the in-person models because the virtual laboratories sparked students’ interest to perform the in-person experiments.

The study by Wang et al. (16) reported that the blended model of physiology laboratory was more effective for students’ motivational learning as compared with the in-person and the standalone virtual laboratories. This is likely because the prelaboratory simulations of the experiments allow the students to practice the procedures, therefore enhancing their confidence to perform the experiments in-person.

On the other hand, the study by Dantas and Kemm (7), which tested a web-based laboratory program that included a virtual physiology laboratory as a subset of the program, showed that the program did not stimulate students’ motivational learning. Due to the design of the program, it could not be determined if the perception of poor motivational learning was attributed to the virtual laboratory component or the entire program. A potential confounding variable is the occurrence of technological advances, including the improvement of user interface and usability of interactive programs, which occurred between 2008 and the other relatively recent studies. As the improvements are expected to enhance the user experience of virtual instruments, motivational learning is likely to be stimulated. Therefore, the absence of motivational learning in this study may be attributed to the lack of realism of the student interface of the virtual physiology laboratory.

As shown, although the studies largely supported that virtual physiology laboratories are effective for students’ motivational learning, the small number of studies assessed and the subjective evaluations lowered the reliability of these evaluations. Thus the evidence is inconclusive in showing that virtual physiology laboratories are effective for students’ motivational learning.

Technical Learning

Compared with the conceptual and motivational learning, a smaller number of studies investigated the effectiveness of virtual physiology laboratories for students’ technical learning. The three studies that evaluated technical learning reported that virtual physiology laboratories were effective for students’ technical learning. This was examined using quantitative and qualitative analysis, based on a mixture of test scores (16, 37) and student perceptions (16, 41). It is worth noting that all studies investigated only the efficacy of virtual physiology laboratories for procedural skills, without testing other aspects of technical learning, such as research skills.

Based on students’ scores, Moreno-Ger et al. reported that the blended model of virtual laboratory was more effective than the in-person laboratory for students’ technical learning (37). After practicing the experiment using the prelaboratory simulation, the students showed a higher level of comprehension of the experimental procedures, and a higher level of precision when performing the experiment in-person. Whereas Wang et al. (16) reported that the blended model, the standalone virtual model, and the in-person model were equally effective for students’ technical learning as all students showed an equal understanding of the experimental equipment and procedures. Based on student perceptions, Durand et al. (41) reported that although the virtual physiology laboratories were as effective as the in-person laboratories in developing students’ technical skills, they differed in terms of the skill sets that they developed. For example, the in-person laboratories developed students’ “hands-on skills” through tactile experiences and the virtual physiology laboratories developed students’ ‘information technology skills through manipulating online software to actively explore physiology concepts. Although all analyzed studies showed that virtual physiology laboratories are effective for students’ technical learning, the small number of studies assessed lowers the generalizability of this finding, therefore, this finding is inconclusive and requires further validations.

Alignment with Past Reviews

Prior systematic reviews have investigated the effectiveness of virtual laboratories across science disciplines such as chemistry and biology (11, 12, 27, 45, 46). A common difficulty for the past and the current reviews is the lack of standard terminology to define the variety of unconventional laboratory classes that are delivered virtually. For instance, a “remote laboratory” has been defined in some studies as being delivered entirely online while others defined it as a hands-on laboratory that can be performed outside of the conventional laboratory environments (27). Moreover, the lack of consistency in the evaluative parameters of student learning further complicates the research findings. For example, studies that supported virtual laboratories over in-person laboratories were largely based on academic performances (assessed using test scores), whereas studies that supported in-person laboratories over virtual laboratories were largely based on student perceptions (assessed using surveys) (45).

One similarity across all systematic reviews of virtual laboratories is that the included studies focused predominantly on the effectiveness of virtual laboratories for students’ conceptual learning, less on motivational learning, and rarely on technical learning (12, 46). Furthermore, the aforementioned reviews found that conceptual learning was examined largely through test scores using the pretest-posttest design or the posttest-only design (quantitative analysis) and, to a lesser extent, through student perceptions (qualitative analysis). Since quantitative analyses are generally associated with fewer experimental biases and a higher power as compared with qualitative analyses, studies that utilized quantitative analyses show higher reliability. As prior reviews largely assessed studies that utilized quantitative analyses, their finding that virtual laboratories are effective for students’ conceptual learning robustly supports this conclusion of the current review.

In comparison, investigations of students’ motivational learning associated with virtual laboratories showed conflicting results due to the large variety of virtual laboratories. According to Lewis (11), the design (e.g., asynchronous/synchronous delivery) of the virtual laboratory plays a significant role in determining student engagement and motivation to explore science concepts. Moreover, Faulconer and Gruss (27) agreed that the virtual laboratories must be designed to be highly engaging to maintain student motivation, an aspect which is absent in many existing virtual laboratory models. In this current systematic review, the designs of virtual physiology laboratories that were shown to motivate students’ learning were simulations (i.e., virtual experiments) and video demonstrations (16, 24, 25, 37, 41), whereas a web-based laboratory course was not associated with student motivational learning (7). Another factor to consider is the synchronicity of the virtual laboratories; however, we found no clear association between synchronicity and student learning. In summary, the design of virtual physiology laboratories is a potential confounding variable that may have caused the variations in students’ perceptions of motivational learning. As a result, despite the largely supportive view of virtual laboratories in effectively stimulating students’ motivational learning, further research into the educational impacts of specific designs of virtual laboratories is needed.

Limitations and Future Implications

The majority of the studies (7 out of 11) for this review investigated virtual physiology laboratories in the blended delivery mode, making it difficult to attribute student learning solely to the virtual component of the laboratory. Furthermore, none of the studies specifically investigated the effectiveness of virtual physiology laboratories for the development of students’ research skills, a crucial element of physiology laboratories. We also found that the use of the search term “undergraduate” led to the exclusion of a couple of papers. However, initial searches with the term “student” were too broad and led to thousands of papers from the preuniversity, or school, sector. Future studies could broaden the coverage of evaluations by using synonyms of “undergraduates” such as “college” and “higher education.” Lastly, all of the studies did not provide detailed descriptions of the laboratory models used, which hindered the potential to provide recommendations for virtual laboratory models. Therefore, future studies should also include detailed reports of laboratory models.

Future studies are warranted to investigate whether specific designs of virtual laboratories support student learning. Further studies are also required to answer the question as to whether virtual laboratories are effective for research skills development. These should use valid and reliable research strategies, such as having large, randomized, and demographically consistent experimental and control groups, with participants assessed by quantitative analyses, high response rates, and rigorous data analysis.

Conclusions

This systematic review has critically evaluated the existing literature to determine the effectiveness of virtual physiology laboratories for student learning (conceptual, motivational, and technical). The findings of this study suggest that virtual physiology laboratories are effective for students’ conceptual learning. However, it is inconclusive as to whether virtual physiology laboratories are effective for motivational and technical learning, with further studies required to investigate this.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

X.Z. and J.C. conceived and designed research; X.Z., D.A.-M, and J.C. performed experiments; X.Z., D.A.-M., and J.C. analyzed data; X.Z. and J.C. interpreted results of experiments; X.Z. and J.C. prepared figures; X.Z., D.A.-M., and J.C. drafted manuscript; X.Z. and J.K.C. edited and revised manuscript; X.Z. and J.C. approved final version of manuscript.